Pacific Northwest forests have undergone substantial changes in the past century as a result of extensive logging, changes in forest management methods, development, and climate change. These forests contain a dense network of small headwater streams that supply cold, clean water and support rich amphibian diversity. The USGS Washington Water Science Center, USGS Northwest Climate Adaptation Science Center, and the Washington Department of Fish and Wildlife are developing models and maps of amphibian habitat in headwater stream ecosystems. We are collaborating with academic, private, tribal, state, and federal partners to ensure our project will produce useful data products for natural resource managers. We facilitate virtual triannual meetings of a Technical Advisory Committee (TAC) through which we provide project updates and receive helpful input in a structured discussion. We seek guidance from the TAC on habitat field survey protocol, the use of spatial datasets to develop proxies for important characteristics of amphibian habitat, and the modeling approach for six target amphibian species in western Washington and northwestern Oregon west of the Cascade Range crest. Spatial datasets include precipitation, stream temperature, streamflow permanence, land cover, land use, and lidar-derived elevation. We are hydrologically conditioning 3-m resolution lidar-derived digital elevation models for accurate hydrologic routing through the landscape and drainage area delineation. We will use estimates of streamflow, stream power, and valley confinement combined with pre-existing datasets for canopy cover and lithology to derive proxies of channel substrate grain size and presence of large instream wood. Random sampling was stratified by EPA Ecoregion to select 10-digit hydrologic unit code watersheds (HUC10s) from the USGS National Hydrography Dataset. In summer 2025, we conducted field surveys of 122 streams across eleven HUC10s, which will be used to validate the habitat proxies derived from spatial datasets. The TAC has indicated that multi-species occupancy and joint species distribution models that include both current and potential habitat refugia for the six species of interest will yield the most useful output for natural resource managers. Through consistent dialogue with the TAC, modeled outputs will be tailored to the needs of regional natural resource managers.